Expand my Community achievements bar.

Your voice matters! Help shape the future of Adobe Journey Optimizer by sharing your feedback.
SOLVED

Dataset lookup in AJO not working

Avatar

Community Advisor

Hi Team,

Recently a new feature was introduced in AJO which allow you to lookup aep data while targeting customers via email. https://experienceleague.adobe.com/en/docs/journey-optimizer/using/content-management/personalizatio...

I created a profile schema and a lookup schema address both enabled for profile plus also created datasets for both(also enabled for profile). Profile is linked to address lookup via a address id. In audience creation, i see that profile is linked to address properly (i can see all fields of address lookup when i select profile attribute which means the relationship is set properly).

 

However in AJO when i send an email to the customer and fetch his address details via the address id , i always get empty address.

 

Did anyone tried this already?

 

Just for information: I have already enabled dataset for lookup as mentioned on the document 

 

Thanks,

Arpan

Topics

Topics help categorize Community content and increase your ability to discover relevant content.

1 Accepted Solution

Avatar

Correct answer by
Employee

Hi @arpan-garg 

There was a JIRA created for this issue, let me also answer here:

Initially - dataset ingestion was re-enabled multiple times, and when ingestion is enabled, then we ingest whole dataset. And at that time there were multiple records for the same testing Id.
And we get a random record having the test id in this case.

Recently I saw few updates to the dataset, for the same testing id. And service returns the latest update, as expected.

If dataset ingestion doesn't get disabled, it should continue working as expected as long as update batches have unique ids.

View solution in original post

16 Replies

Avatar

Employee Advisor

HI @arpan-garg -

Couple of questions. 

 

1) Does your schema have "addressID" set as the primary identity? 

2) Do you have unique "addressID" for each row in the dataset? 

3) When you used the API call to enable your dataset for lookup did you get a Success 200 message returned? 

4) When using the function in personalization editor did you get any validation errors? 

5) Did you make sure to use the full path from the schema when referencing the fields of data you wish to bring in? 

6) You can set the dataset look expression to be required="true" which should result in an exclusion if a match is not found.

7) Did you try putting in a static key into the expression to see if you can just get a match to work? 

 

Make sure that if you are using a dynamic attribute for the ID that you remove the quotes. i.e. id=profile.customer.addressID. if you have id="profile.customer.addressID" it will look for this static value and find no results. 

Avatar

Community Advisor

@Ivan_Mironchuk ! Your last suggestion, to avoid surrounding the id with quotes, helped me resolve the issue—I was making the same mistake. However, I would greatly appreciate it if Adobe could address this inconsistency in their documentation. In this guide, it is still mentioned to use quotes.

 

Regarding your second point about ensuring only unique entry exists for each address ID, I encountered a scenario where a customer's address changed from Texas to Arizona. However, the lookup always returns the old address (Texas). How can this situation be managed? Lookup functionality will only be truly useful if it fetches the latest data from the dataset, rather than the first record.

 

What do you suggest?

Avatar

Employee Advisor

Can you DM me with further details on how you are replacing the data in the dataset? 

Avatar

Community Advisor

Sent

Avatar

Level 2

We are having the same issue. Unable to replace the data that's in the dataset. Only works if I delete the dataset, recreate it, and re-ingest. Any solve here?

Avatar

Employee

@mm2021, @arpan-garg Can you share your datasetId in a DM?

Avatar

Community Advisor

@Arnas_Valeika  - 6759606d0561a02aef3d10aa 

Avatar

Correct answer by
Employee

Hi @arpan-garg 

There was a JIRA created for this issue, let me also answer here:

Initially - dataset ingestion was re-enabled multiple times, and when ingestion is enabled, then we ingest whole dataset. And at that time there were multiple records for the same testing Id.
And we get a random record having the test id in this case.

Recently I saw few updates to the dataset, for the same testing id. And service returns the latest update, as expected.

If dataset ingestion doesn't get disabled, it should continue working as expected as long as update batches have unique ids.

Avatar

Community Advisor

Hi @Arnas_Valeika  - It works now. Good to know the logic on how it works, no where in the documentation was it mentioned that we should not renable it for lookup multiple times and since this functionality is new , i agree we did it multiple times to test

Avatar

Community Advisor

@Ivan_Mironchuk  - One more question, does the functionality of data set lookup actually require a relationship to be set between the profile and address. Or will this work without even establishing a relationship?

Avatar

Employee Advisor

No relationship required. Dataset does not need to be enabled for profile. 

Avatar

Community Advisor

@Ivan_Mironchuk  If in lookup dataset a certain primary key has a new value. For example details of a particular address id changed due to renaming of a street then is there a way that the dataset lookup feature can pick the latest entry for that id instead of always picking up the first value in the lookup table?

 

Avatar

Level 4

Facing the same issue for my dataset lookup as well. Any solution, @arpan-garg @Ivan_Mironchuk ?

Avatar

Administrator

Hi @arpan-garg

Can you please share the solution that Ivan helped you achieve with the community?

Thanks!



Sukrity Wadhwa

Avatar

Community Advisor

Hi @Sukrity_Wadhwa  - Haven't got any response on this from @Ivan_Mironchuk , still waiting.

Avatar

Employee Advisor

@arpan-garg @Will need to get a support ticket raised to investigate the details. How are you loading the new batches of data? The revised data should be available for lookup once the batch is completed processing